SGD drives changes in fish community structure and function

Who is there - Community Composition

NMDS plot of reef fish species abundance community composition (n = 3 / tide / SGD influence / habitat type; stress = .15). Data underwent Hellinger transformation - good for species abundance data, gives low weights to variables with many zeros.

Reef & sand habitats have different communities of fish species

There is a significant difference between fish assemblages in sand habitats and reef habitats (ANOSIM, R = .7025,p =.0001)

Communities in sand habitats are more variable than in reef habitats

Communities in sand habitats are significantly more variable than those in reef habitats (ANOVA / betadisp; F = 17.71, p = .0003)

Variability is driven by groundwater influence, which creates different fish communities in sand habitats but not in reef habitats

In sand habitats, there are significant difference in fish assemblages between high and low groundwater influence (ANOSIM, R = .4204, p = .0038).

Species Community Plots

Questions

ANOSIM tests whether two or more groups of samples are significantly different.
Do significant differences between habitat types mean they can be analyzed separately can only specify one grouping (I think) & cannot do interactive effects, how to look at interaction between SGD & tide?
How to track changes between high and low tide at one site? calculate distance between points and then group distances?

Taxonomic family groupings

Reef & sand habitats have different communities of fish families

A significant difference between fish families in sand habitats and reef habitats (ANOSIM, R: .5667, p=.002)
Kuhliidae are a family of flagtails: most are euryhaline and often found in brackish water, but the genus also includes species restricted to marine or fresh water.

Across all habitat types, groundwater influence drives difference in communities of fish families

A significant difference between fish families in high groundwater vs low groundwater (ANOSIM, R: .1375, p=.0067 ) Not the case for species differences, only families.

Families in sand habitatsare more variable than in reef habitats

Communities in sand habitats are significantly more variable than those in reef habitats (ANOVA / betadisp; F = 49.91, p < .00001)

Variability is driven by groundwater influence, which creates different fish communities in sand habitats but not in reef habitats

In sand habitats, there are significant difference in fish assemblages between high and low groundwater influence (ANOSIM, R: .5667, p=.0019)

Broad divesity stats

Species richness is higher in reefs than sand habitats (3 way ANOVA, F = 17.883, p = .006)

What are they doing? Feeding analysis

Tide and groundwater significantly affect bite rates on different substrate types

Significant three way SGD:tide:substrate interaction.

No significant relationships because standard deviation too high. Slight SGD influence (p = .07)

A diverse set of species feed on plankton in groundwater!

33 species that have observations feeding on plankton in high SGD regions at low tide

Dietary Plasticity

Next Steps

Analyze two more days of footage

Connect these results to herbivory choice experiment

Higher rates of herbivory on reefs not necessarily due to differences in communities, more due to feeding rates. In sand habitats could be due to different communities.